15:I[981,["511","static/chunks/511-fc0d2f24caf5741b.js","183","static/chunks/183-6beeeb801c274aaa.js","474","static/chunks/app/%5Btype%5D/layout-347af3707a752b1a.js"],"Posts"] 16:T440,Hey HN, I built a Chrome extension called Legal Eyes that rewrites your casual messages into sharp, professional legal language.

Use cases: - Freelancers chasing late invoices - Founders dealing with vague contract terms - Anyone trying to sound more serious or formal (without being a lawyer)

No prompt engineering, no copy-pasting. It sits in your browser, and converts selected text with a click. I built it using React, Supabase (for auth and usage tracking), OpenAI (for the rewriting), and deployed via Vercel.

I launched this to help the little guys when they need a stronger voice. I hope this goes out into the world and helps someone :)

Would love your feedback on: - Features you'd expect in v2 - Thoughts on monetizing vs keeping it free - Weird or niche use cases I haven’t thought of

You can try it here (free version available): [https://legaleyes.uk](https://legaleyes.uk)

Thanks! Happy to answer anything.17:T5c9,So I vibe-code a lot these days and recently i decided to give the same prompt to several llms, then get their codes and later give each code to every single one of them to ask which one they think is the most useful without telling them that they or the other 2 llms wrote it. The overall consensus is: gpt5. True I only compared gpt5 vs claude 4.1 vs qwen 230bn. OSS 120b, gemini and grok 4 were excluded since well i don't have the time. And obvious failures like amazon nova or anything from meta weren't even planned. Deepseek (both) seem a bit underperforming . Personally I'd say it's a close call between claude opus 4.1 vs both gpt4 and gpt5 (ironically gpt5 sometimes performs worse than 4, i think this has been addressed by many people

Lemonade: Local LLM Serving with GPU and NPU Acceleration